The Signal
BTQ Brings MIMIQ to qBraid’s 35,000 Developers
BTQ Technologies has made QPerfect’s MIMIQ quantum emulator available through qBraid’s managed API, reaching a community of more than 35,000 developers. The integration supports large-circuit testing with Matrix Product State methods, subje
Quantum.world Team · Written with AI · fact-checked ·
“By bringing MIMIQ to qBraid's developer community, we're expanding access to large-scale quantum emulation and supporting the development of the broader quantum computing ecosystem.”
Olivier Roussy Newton, CEO of BTQ Technologies
BTQ Technologies announced on October 29 that developers can run its MIMIQ quantum emulator through qBraid’s managed API. MIMIQ was developed by QPerfect, BTQ’s wholly owned subsidiary. The integration gives qBraid’s community of more than 35,000 developers a new option for testing quantum circuits without setting up MIMIQ infrastructure separately.
The announcement puts the emphasis on access to software for work that can be difficult to test on quantum processing units (QPUs). It does not report how many qBraid users have tried MIMIQ, what the service costs, or what computing resources users receive. BTQ describes the integration as available now; the release does not publish adoption or performance figures from the new service.
Circuits run through qBraid Runtime
Users submit circuits through qBraid Runtime and its managed API. The connector, called a Provider Plugin, is open source under the Apache License 2.0 and builds on QPerfect’s open-source mimiqcircuits-python library. The workflow is intended to spare users from configuring MIMIQ infrastructure themselves.
BTQ and qBraid describe several uses: testing algorithms at sizes closer to those planned for hardware, examining noise, and exploring error-correction protocols. Such simulations let researchers investigate circuits before spending time on a QPU, but they do not replace hardware runs. Results from an emulator alone cannot establish how a physical processor will perform.
MPS extends some simulations, not every circuit
Conventional state-vector simulation requires memory that grows exponentially with qubit count: each added qubit doubles the state representation. MIMIQ combines optimized state-vector emulation with Matrix Product State (MPS) techniques, which represent quantum states in compressed form. That compression can make much larger simulations practical when a circuit’s structure and entanglement suit the method.
BTQ says MIMIQ can simulate circuits with thousands of qubits and millions of gates under those conditions. The caveat matters: MPS does not remove the computational cost for every circuit, and high entanglement can limit how much compression helps. The announcement also refers to controlled truncation error for circuits beyond exact state-vector simulation, so users must consider approximation when interpreting results.
A study cited in the release, “Comparative Benchmarking of Utility-Scale Quantum Emulators”, tested seven emulators on 13 benchmark circuits ranging from 4 to 1,024 qubits. Its authors report that MPS-based emulators solved eight benchmarks at the maximum tested size and 12 at 100 qubits or more in under five minutes. Those are study-wide findings, not a performance guarantee for every MIMIQ workload or for the qBraid integration.
Access expands beyond specialist installations
qBraid says its platform serves developers in more than 100 countries and at more than 2,000 universities and research institutions, as well as national laboratories and companies. The platform has also served for five consecutive years as the official development platform for MIT’s iQuHACK quantum hackathon, according to BTQ’s release. Those figures describe qBraid’s reach; they do not indicate how many users will run MIMIQ.
The practical change is a route to MIMIQ within qBraid Runtime, alongside an openly licensed connector and no separate MIMIQ infrastructure setup. For developers, the test is whether that route makes it easier to examine a given circuit before a hardware run—and whether the circuit’s structure lets MPS simulation scale.